From War to Peace on the Mozambique-Malawi Borderland 9780585443874

Citation preview

International African Library 26 General Editors:]. D. Y. Peel, Colin Murray and Suzette Heald

FROM WAR TO PEACE ON THE MOZAMBIQUE-MALAWI BORDERLAND

The International African Library is a major monograph series from the International Mrican Institute and complements its quarterly periodical Africa, the premier journal in the field of Mrican studies. Theoretically informed ethnographies, studies of social relations 'on the ground' which are sensitive to local cultural forms, have long been central to the Institute's publications programme. The IAL maintains this strength but extends it into new areas of contemporary concern, both practical and intellectual. It includes works focused on problems of development, especially on the linkages between the local and national levels of society; studies along the interface between the social and environmental sciences; and historical studies, especially those of a social, cultural or interdisciplinary character.

International African Library General Editors

}. D. Y. Peel, Colin Murray and Suzette Heald Titles in the series published in the United Kingdom by Edinburgh University Press: 1 Sandra T. Barnes Patrons and power: creating a political community in metropolitan Lagos 2 Jane I. Guyer (ed.) Feeding African cities: essays in social history 3 Paul Spencer The Maasai of Matapato: a study of rituals of rebellion 4 Johan Pottier Migrants no more: settlement and survival in Mambwe villages, Zambia 5 Gunther Schlee Identities on the move: clanship and pastoralism in northern Kenya 6 Suzette Heald Controlling anger: the sociology of Gisu violence 7 Karin Barber I could speak until tomorrow: oriki, women and the past in a Yoruba town 8 Richard Fardon Between God, the dead and the wild: Chamba interpretations of religion and ritual 9 Richard Werbner Tears of the dead: the social biography of an African family 10 Colin Murray Black Mountain: land, class and power in the eastern Orange Free State, 1880s to 1980s 11 J. S. Eades Strangers and traders: Yoruba migrants, markets and the state in northern Ghana 12 Isaac Ncube Mazonde Ranching and enterprise in eastern Botswana: a case study of black and white farmers 13 Melissa Leach Rainforest relations: gender and resource use among the Mende of Gola, Sierra Leone 14 Tom Forrest The advance of African capital: the growth of Nigerian private enterprise 15 C. Bawa Yamba Permanent pilgrims: the role of pilgrimage in the lives of West African Muslims in Sudan 16 Graham Furniss Poetry, prose and popular culture in Rausa 17 Philip Burnham The politics of cultural difference in northern Cameroon 18 Jane I. Guyer An African niche economy: farming to feed Ibadan, 1968-88 19 A. Fiona D. Mackenzie Land, ecology and resistance in Kenya, 1880-1952 20 David Maxwell Christians and chiefs in Zimbabwe: a social history of the Hwesa people c. 1870s-1990s 21 Birgit Meyer Translating the devil: religion and modernity among the Ewe in Ghana 22 Deborah James Songs of the women migrants: performance and identity in South Africa 23 Christopher 0. Davis Death in abeyance: illness and therapy among the Tabwa of Central Africa 24 Janet Bujra Serving Class: masculinity and the feminisation of domestic service in Tanzania 25 T. C. McCaskie Asante identities: history and modernity in an African village 26 Harri Englund From war to peace on the Mozambique-Malawi borderland

FROM WAR TO PEACE ON THE MOZAMBIQUEMALAWI BORDERLAND HARRI ENGLUND

EDINBURGH UNIVERSITY PRESS for the International African Institute, London

© Harri Englund, 2002 Edinburgh University Press Ltd 22 George Square, Edinburgh Transferred to digital print 2006 Typeset in Plantin by Koinonia, Bury, and Printed and bound by CPI Antony Rowe, Eastbourne A CIP record for this book is available from the British Library ISBN-10 0 7486 1577 6 (paperback) ISBN-13 9 7807 4861577 3 (paperback) The right of Harri Englund to be identified as author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988.

CONTEN TS

List of Maps, Tables and Figures Preface

Introduction

Vl Vll

1

1 Borders Drawn, Borders Crossed

37

2 The Paths to War

58

3 Refugees from Afar

81

Gendered Exile

99

4

5 Migrants amongst Refugees

115

6 Paradoxes of Repatriation

137

7 Value, Power and 'Social Capital'

160

Epilogue: Borderland Revisited

170

Appendix Notes Bibliography Index

186 189 196 212

LIST OF MAPS, TABLES AND FIGURES

Maps 1 Malawi and its neighbours 2 The Dedza-Ang6nia borderland 3 The divisions of Chitima village 4 Towns and villages of the Dedza-Ang6nia borderland Tables 1.1 Investment in livestock among local and refugee households in Mfuno village, 1992 1.2 Female-headed households and the investment in livestock among locals and refugees in Mfuno village, 1992 Figures 2.1 Achingano's wives and children who migrated from Domwe 2.2 Ferdinand and persons in Jeremiya's group 2.3 Key actors in Chitima politics 3.1 The adult refugees from Lumbe in Mfuno 3.2 Nachisale's cross-cousin and Ben's siblings 3.3 Mbewe's children and heir; John and Filipo 4.1 Nasoweka's siblings 4.2 Nantuwa's family 4.3 Namanyada, Malitina and Rose 5.1 Joji and Martha 6.1 Sadlek and the founders of Mtunda division 6.2 Alfredo's and Sandikonda's pedigree and conjugal bonds Chart I Genealogical relations among Chitima and Mfuno villagers Chart II Genealogical relations among Chitima and Mfuno villagers

xi

xn xn xm

54 55 60 61 64 85 88 93 103 107 109 129 152 156 187 188

PREFACE

The war in postcolonial Mozambique enjoys dubious fame among the political conflicts of the late twentieth century. Up to one million persons died- some directly of warfare and others of war-related disease and famine - one-and-half million fled to neighbouring countries, and some three million were displaced within Mozambique. It was not like the 1994 genocide in Rwanda, a sudden carnage that stunned the world mass media for a few terrible, passing moments. Waged from the late 1970s until1992, the war in Mozambique had its atrocities spread over several years, often in remote rural areas. Only occasional media coverage publicised the war outside Mozambique, and humanitarian interventions consisted mainly in relief aid to refugees who had crossed international borders. And yet, many of these refugees were able to relate stories which insisted, in haunting detail, that all evil had taken over Mozambique. Renamo rebels, fighting the Frelimo government, often gave a face to this evil. Consider one common story from northern Mozambique, recounted by refugees in Malawi and Zambia. Machanga, as Renamo was locally known, came to a village, looking for loot and food. The guerrillas found a woman, carrying an infant on her back, alone at her house. They ordered her to prepare food. She replied, in anguish, that there was no relish in the house. The guerrillas had become used to having meat at every meal and could not accept the woman's excuses. They seized her baby, killed it with a machete, and roasted it outside the house. According to one version of the story, the mother was forced to taste her baby's flesh. According to another version, she was herself killed when the machete slashed in one terrible stroke both the baby and the mother carrying it. What is there to be said, by way of analytical ethnography, about people who lived with such evil in their midst? Anthropologists, responding to the moral and intellectual need to confront the atrocities of our times, are debating ways in which ethnographic accounts could be written against violence (see e.g. Taussig 1987; Daniel 1996; Nordstrom 1997; Zur 1998; Besteman 1999). Much of this debate concerns the representation of violence. It brings into a critical focus, on the one hand, the sanitising of violence which erases all that is disturbing about violence. On the other hand, it also

viii

FROM WAR TO PEACE

questions the representation of violence as a bizarre ethnographic fact. The ethnographic representation of violence, however, is not the main concern of this book. Although violence and terror assume self-evident importance as topics of ethnographic inquiry in some conflicts, the focus on violence may also privilege the dramatic and the extraordinary, thereby committing the old sins of exoticism. In the case of Mozambique, evidence on mutilated war victims, massacres and the forced recruitment of childsoldiers must be put in perspective. This book is a study of villagers who avoided such atrocities. They certainly had more than their fair share of violence, but a profoundly distorted picture of the Mozambican war would emerge if violence were thought to have devastated the country with the same intensity everywhere. The fact that stories about extreme violence had spread widely with flows of refugees did not mean that everyone was obsessed with such stories. Few villagers who appear in this study felt any special urge to recount stories such as the one above, probably because few had personal testimony to deliver. Villagers were outspoken in their criticism of the parties of the war, but their criticism often revolved around the mundane minutiae of strained personal relationships. The complex moral underpinnings of those relationships form the subject-matter of this book. The subject-matter evokes, in other words, a whole range of issues other than the ethnographic representation of violence. Personal relationships provide a standpoint from which to address other relations of varying scale. The war and displacement which the villagers in the case studies of this book endured were consequences of the processes of state formation, processes which were themselves shaped by the transnational pursuit of political influence and material accumulation (cf. Tilly 1985; Ferguson 1990). Although very different in scope and achievement, two recent anthropological studies of conflict in Africa admirably describe the interplay between the lived experience and large-scale processes. One qualifies the apparently anarchic violence of youths in Sierra Leone by drawing attention to the declining neopatrimonial state (Richards 1996). The other puts the approaching civil war in southern Sudan in the context of existential uncertainties, ranging from relations between old and young, women and men, to the meaning of death and divinity (Hutchinson 1996). In these studies, a 'context' does not make violence seem less disturbing. It introduces, rather, nuance and complexity into ethnography. This book brings ethnographic analysis to bear upon war and exile in the borderland of Mozambique's Ang6nia District in Tete Province and Malawi's Dedza District in Central Region. Based on archival research, eighteen months' fieldwork in 1992-3 and revisits in 1996-7, this study examines the paths of particular villages to nationalism, war, displacement, repatriation and post-war uncertainty. The focus is on Dedza-Ang6nia

PREFACE

ix

borderland villagers' relationships; how these relationships variously enabled and constrained villagers in their aspirations during and after the war, and in their pursuit of refuge during the displacement. A major theoretical-cummoral assault is made on naturalism, which undermines complexity by seeking determinate patterns of social life on the basis of kinship systems, ethnic loyalties, state ideologies, universal refugee experiences, and so on. Throughout the book, a key question is the extent to which internal social forms are to be understood in the light of external interventions - and what indeed could be the best way to imagine the external-internal divide, deceptively obvious as it is in villagers' encounters with revolutionary and counter-revolutionary movements. Given the importance of fieldwork for the contribution which this book aspires to make, the Dedza-Ang6nia borderland must be acknowledged as an intellectual setting on a par with the academic institutions where the writing of this book has taken place. Through their arguments and aspirations, conflicts and compassion, villagers in the Dedza-Ang6nia borderland continuously challenged my preconceptions of what their lives amid crises might be like. Their views and acts deconstructed the categories that were at my disposal- 'victims', 'hosts', 'refugees' -in many complex ways. It is, therefore, with some dismay that I safeguard their privacy by using pseudonyms in the case studies. It is surely awkward to acknowledge the help and inspiration one has received from one's friends and interlocutors by using pseudonyms for them! And yet, perhaps academic treatises are not the places to repay the kinds of debts I have incurred in the Dedza-Ang6nia borderland. Those debts are to be readjusted by continuing the long conversation I have become engaged in. In addition to the Dedza-Ang6nia borderland, the Department of Social Anthropology at the University of Manchester has been another important intellectual setting during the gestation of this book. I am grateful to the Academy of Finland for funding both my study in Manchester and my fieldwork in Malawi and Mozambique. The Leach/Royal Anthropological Institute Fellowship in 1997-8 brought me back to Manchester, and gave me an opportunity to embark on revision and rewriting in a supportive environment. The Nordic Africa Institute co-funded my revisit to Malawi and Mozambique in 1996-7, and kindly enabled me to take advantage of the Fellowship by granting me leave of absence. Although my intellectual debts to the Manchester School had already begun to accumulate before my first arrival in Manchester in 1991, two anthropologists, in particular, have taught me to work in a critical relation to its substantial ethnographic and analytical legacy. One of them is Richard Werbner, whose faith in my project has been unfailing even during the most trying times. Friendship and intellectual curiosity have always combined in

X

FROM WAR TO PEACE

his critical response, extending far beyond the formal obligations of a 'research supervisor'. The other is Bruce Kapferer, whose commitment to anthropology has been a source of inspiration. Several other friends and colleagues assisted me in my research and in the making of this book. Many made their contributions during various phases, possibly even without realising it. I must mention at least Jocelyn Alexander, Alice Dinerman, John Gledhill, Maia Green, Barbara Harrell-Bond, Penelope Harvey, Tim Ingold, James Leach, David Maxwell, JoAnn McGregor, Heike Schmidt, Matthew Schoffeleers, Anthony Simpson, Marilyn Strathern, Megan Vaughan and K. B. Wilson. Philip Gillon and the editors and anonymous readers of the International African Library series had the unenviable task of distilling a book from my writings. Although their contribution has been profound, all the fallacies and weaknesses remain my own. The Centre for Social Research at the University of Malawi provided the institutional affiliation during my fieldwork. My discussions with scholars in different departments of that university - such as Wycliffe Chilowa, Karin Hyde, Paul Kishindo, Kings M. Phiri and E. S. Timpunza-Mvula - were important to my understanding of Malawi. Charles Malunga facilitated my research in the National Archives of Malawi. Finally, I owe a special debt of gratitude to Charles Namondwe, my mchimwene and confidant.

Map 1 Malawi and its neighbours • town

.A. mountain 100

0 km

TANZANIA

ZAMBIA

• Kasungu MALAWI

MOZAMBIQUE

···~ ...

,._Tete ', :....

Map 2 The Dedza-Ang6nia borderland

•

• Kanyama

•

• • • •

-main road - - international boundary • village

2

0

•

km

Map 3 The divisions of Chitima village

Manguluwe

-

international boundary

0.5

0 km

Maphiri Machanza

Map 4 Towns and villages of the Dedza-Ang6nia borderland

.

• Bembeke

•

A

Domwe

Lumbe

Ntcheu

main road international boundary • • 6.

town village mountain

0

10 km

INTRODUCTION

The peace treaty that ended the prolonged civil war in Mozambique was signed in 1992. On the first anniversary of the signing of the treaty in 1993, a moment of sheer euphoria took place in Villa Ulongwe, the district capital of Ang6nia in Tete Province. Folk dances were performed, senior district and provincial officials made speeches, and business in the beer-halls of the town boomed as jubilant crowds celebrated until the early hours of the morning. Peace found its greatest confirmation in the mirthful hullabaloo which defied the darkest hours. Insecurity no longer forced people to retreat indoors at dusk, fearful of the nocturnal attacks by guerrillas and unknown marauders. A whole year had shown that peace could endure in the troubled land, Mozambique. The speeches at the celebration lauded a new Mozambique, a nation united in performing the tasks of reconstruction and reconciliation. Three men from the remote border village of Chitima participated in the anniversary party in Villa Ulongwe. They were: Rafaelo, the village headman; Luis, Rafaelo's full brother and the village secretary of the ruling Frelimo party; and Alfredo, Luis's henchman and a Frelimo activist. For all three of them, the attainment of peace was also a personal victory. After seven years of exile in Malawi, they had recently witnessed the reconstruction of their village in Mozambique. Its leadership was again firmly in the hands of the three men. There appeared to be reason for them to celebrate, indulging like sophisticated townsmen in drinking bottled beer instead of the locally brewed opaque brand. Drunk as lords, they were not deterred by the thought of the recent census which they had organised in the village. It was the money they had collected during the census that financed their indulgence. Guilt was to come later, prompted by villagers' criticism of the money wasted on the party and vociferous questioning about whether the border village would ever be integrated into the new Mozambique. Another snapshot, some seven years earlier in 1986, presented a different picture. Luis and Alfredo were balanced on their knees inside a small enclosure. They were surrounded by half-a-dozen goats. Outside the enclosure

2

FROM WAR TO PEACE

stood four guards belonging to the Renamo rebel movement. Their faces were hidden behind pieces of cloth. Their hands clutched AK 4 7 rifles. The day had begun with Renamo guerrillas breaking into the houses of the two men at dawn. Although this ominous invasion did not lead to murders, as many had feared, Luis and Alfredo suffered tribulations that were harrowing enough. From dawn to dusk, they were forced to remain on their knees. To make their humiliation complete, the goats were kept in the enclosure, tormenting them with bleating and excrement throughout the whole long day. When Luis and Alfredo were allowed at last to return to their homes, they could hardly walk. The goats belonged to Rafaelo. When the two men's fate became known, nobody doubted Rafaelo's complicity in their detention by the Renamo guerrillas. His interest in Renamo, the new guerrilla movement, had ceased to be a secret, and his dislike of the Frelimo government was widely regarded as a consequence of his close association with the Portuguese colonial authorities before Mozambique attained independence in 1975. During the colonial period, many nationalists had been harassed as a result of information supplied to the authorities by Rafaelo. When Mozambique became independent, Rafaelo was deposed from his position as headman, and Luis, his younger brother and a pioneer nationalist in the area, assumed the leadership of the village. For Rafaelo and some other villagers, Renamo carried the promise of a new dawn. Their first aim was to suppress Frelimo in the village, and, by humiliating Luis and Alfredo, the new movement flexed its muscles. This book fleshes out the snapshots. Between the impending war of 1986 and the apparent reconciliation of 1993 many events took place, both in the villagers' own lives and in the wider world. Nor, as indicated above, did the guerrillas enter a pristine community when they reached the border village of Chitima in 1986. Colonial and postcolonial interventions had profoundly affected the historical trajectory of the village. For the war in Chitima village, this book discloses a range of underlying processes which contradict two deceptively plausible perspectives: one on the 'obvious' conflict between Frelimo and 'traditional' leaders, and the other on the 'natural' rivalry between the two brothers. In a similar vein, displacement, repatriation and, in the Epilogue, post-war political pluralism and economic liberalisation appear as complex processes in border villages, to be understood with intimate knowledge of villagers' relationships. For analytical ethnography, therefore, the challenge is to study the interplay between 'external' and 'internal' factors; whether, as Jean-Fran~ois Bayart has argued in the study of the state in Africa, 'the production of "internal dynamics" is indissoluble from the interference of "outside dynamics"' (1993: 266).

INTRODUCTION

3

EXTERNAL AGGRESSION, INTERNAL DISCONTENT

No easy consensus seems possible on the causes and nature of the war between Frelimo and Renamo. 1 Waged for most of the years of Mozambique's short history as an independent nation, the war has been subject to various explanatory frameworks, with contradictory perspectives often betraying profound ideological differences. The fundamental dividing line is the question of whether this was a civil war or a national tragedy created by external aggression. In their crudest forms, both perspectives deal with culpability instead of analysing the nuances involved. The civil war thesis blames Frelimo's revolutionary agenda, almost wholly ignoring the hostile international context in which its leadership sought to bring about social and economic transformations (see Hoile 1994). The external aggression thesis, on the other hand, portrays Renamo as a puppet movement designed to 'destabilise' the young socialist state (see Fauvet 1984; Metz 1986; Cammack 1988; Morgan 1990: 605-7; Africa Watch 1992: 17-18, 25-8). Such external enemies included the white minority governments in southern Africa and their right-wing sympathisers in Europe and the United States. The 'paradigm shift' that took place in the studies of the Mozambican war in the late 1980s discovered some wisdom in both perspectives, which, if taken separately, are now seen as fatally inadequate (cf. Clarence-Smith 1989). No serious student of the Mozambican experience doubts that Renamo emerged and advanced by means of substantial foreign military and material aid. The genuinely contentious issue appears to be what Renamo became in the course of the war, and how that transformation took place. The Frelimo government's unpopular resettlement policies and antagonism to religion and 'traditional' authorities have been increasingly researched by scholars as flawed policies which paved the way for Renamo's support within Mozambique (see Geffray 1990; Hall 1990: 59; Morgan 1990; Young 1990; Vines 1991: 93; Simpson 1993: 323-31). From the outset of the war, Frelimo and Renamo plunged into a process in which 'external' and 'internal' factors did not divide into neatly separate sets of causes. Any account of the war that leaves out the history of colonialism in Mozambique, and the waging of the Cold War world-wide, must face the charge of distortion. The international context of the conflict, although often recounted in the literature, merits a brief resume because of the light it sheds on the war as a tragedy that had a history beyond the immediate failures of a new government. Colonialism and the Cold War also merit attention for what they do not explain. Later in this Introduction I review local studies of the war. These studies demonstrate profound variations in the issues and social relationships through which the war was experienced in particular settings. Such local variations show how colonialism and the Cold War, even the activities of the Frelimo government, fall short of providing comprehensive accounts

4

FROM WAR TO PEACE

of the war. The challenge remains ofhow to devise a perspective which makes the appreciation of such variations possible. The same challenge applies in the study of displacement and repatriation. After reviewing local studies, I situate this book vis-a-vis refugee studies and the anthropology of borderlands. My approach is at variance with attempts to theorise a generalised 'refugee experience' or a 'border culture'. The subjects of this study went into exile as persons enmeshed in relationships, and the processes by which they experienced their displacement were indeterminate because of the myriad relationships which they carried with them. Their relationships were shaped, as both before and after displacement, by historical contingencies. The refugee status and the international border certainly appeared in villagers' arguments, but they were, among many other available notions, means to assess and contest relationships as events unfolded. The aim of this book is not, however, to celebrate indeterminacy at the expense of understanding the emergence of determinate, and yet intermittent, power relations. I discuss these analytical concerns and my method of ethnographic representation towards the end of this Introduction. I end with a synopsis of the chapters that follow.

Colonial peculiarities, nationalist struggles Unlike the case in Angola, Portugal's other vast colony in southern Mrica, no portion of Mozambique remained occupied by the Portuguese for all the five centuries that began in the early 1500s with the first permanent Portuguese settlement and ended with independence in 1975 (Marcum 1969; Newitt 1995). Despite their early involvement in the slave and ivory trades, the Portuguese effectively occupied Mozambique only at the end of the nineteenth century, largely to guard the territory against Britain and Germany. Afterwards, Portuguese colonialism in Mozambique retained a distinctive character. The collection of head tax and the migration of male labour to the South Mrican mines were by no means unique in the region's colonial history. More peculiar was the leasing of large parts of territory to foreign-owned companies. In 1891, for example, one-third of Mozambique was controlled by two companies as virtually private colonies. Outside the plantations producing sugar, tea, sisal and copra, commercial agriculture among smallholders took the form of forced cultivation of certain crops, notably cotton, at fixed prices (Pitcher 1993; Isaacman 1996). The Salazar government, which took power in Portugal in 1930, had a clearer vision of empire-building than its predecessors and encouraged Portuguese settlement. Whereas in 1930 the white population of Mozambique was 18,000, by 1960 it had risen to 85,000 (Minter 1994: 14). This was the period of the Portuguese 'shopkeeper colonialism' (Hanlon 1984: 188), with settlers managing foreign companies or their own small-scale commercial endeavours in both urban and rural areas. By 1970 Mozam-

INTRODUCTION

5

bique was one of the most industrialised countries in Africa, but much of its economy and labour force continued to serve its neighbours. The Portuguese settlers themselves hardly feasted on Mozambique's wealth; many were illiterate or poorly educated, and were engaged in menial or semi-skilled tasks as low-level government clerks, taxi drivers and shopkeepers. Black Mozambicans were systematically excluded from having private businesses, and smallholder agriculture was shaped more by coercion than by market incentives. A case study of a borderland village shows how the shopkeeper colonialism could, in a particular setting, become part of the local social fabric. From the viewpoint of the African elite in Maputo (then Louren~o Marques), however, discriminatory practices were unmistakable. By virtue of education, income and life-style, a small number of Mricans were granted an assimilado status, becoming Portuguese in the eyes of the colonial masters (Minter 1994: 14). As the numbers of whites in Mozambique increased, it became clear that such a status secured little more than a position as secondclass, or even third-class, Portuguese. Whites from Portugal, even when poorly educated, were favoured by colonial authorities and employers over mestifos and locally hom whites, not to mention Africans. Until the 1920s, Mozambique witnessed a series of revolts against colonial labour regimes, particularly in the Zambezi valley region (discussed later in this Introduction). A coherent nationalist movement, however, did not emerge until the early 1960s. Apart from short-lived attempts to establish rival nationalist movements in Zambezia, and occasional announcements by exiled Mozambicans in various African capitals, Frelimo encompassed the entire spectrum ofliberation forces in Mozambique. Its origin was in the assimilado and mestifo elite who had been educated by Protestant missionaries from the United States and Switzerland in southern Mozambique, near the capital (see Cruz e Silva 1998). The founding congress took place in Dar es Salaam in 1962. For the fascist government that had assumed power in Portugal in the 1950s, independent Mozambique was not even an option, and the Portuguese security police made it extremely difficult to promote the nationalist cause inside Mozambique (Birmingham 1992: 534). In 1964, the year Malawi gained independence from Britain, Frelimo began to pursue independence through armed struggle. With the aid of bases in Tanzania, and later in Zambia, Frelimo launched attacks in northern Mozambique, achieving 'liberated zones' mainly along Lake Malawi and the Tanzanian border in Niassa and Cabo Delgado Provinces, and in Tete Province. Frelimo's health and education programmes attracted support from the Nordic countries and religious organisations, while the leadership's careful navigation through the Sino-Soviet dispute ensured military aid from both China and the Soviet Union (Minter 1994: 15).

6

FROM WAR TO PEACE

In 1966-9, Frelimo experienced a leadership crisis, not least because of Portuguese efforts to manipulate internal discontent (Minter 1994: 91-2). Although it is patently misleading to represent Frelimo, even during these early years, as a movement of educated southerners (cf. Newitt 1995: 53841; Hall and Young 1997: 17-18), the top leadership was composed disproportionately of those who had benefited from educational advantages in the south. The southern bias was based on class rather than ethnic identity, with mestifOS and whites also assuming important positions in Frelimo. Racialism and regionalism became central sources of concern for the Frelimo leadership, and after the assassination of Eduardo Mondlane, its first president, in 1969, a new sense of common purpose was embodied by Samora Machel. The revolutionary agenda of social justice disarmed many critics in moral terms, but certain divisions prevailed in a latent form, reappearing in the post-independence period. Prominent among these was the association of Makonde people in the north with Frelimo, and the virtual absence of Macua-Lomwe-speaking peoples from Nampula and Zambezia from the top leadership. On the other hand, those educated in the Catholic mission in Beira often felt disadvantaged by Protestant-educated southerners (Abrahamsson and Nilsson 1995: 25-6). Independence in 197 5 owed more to a coup in Lisbon in April 197 4 than to the liberation war in Mozambique - most areas, including cities, had been unaffected by the war. Frelimo thus came to power somewhat unexpectedly, without an election or a referendum. It had only a rudimentary organisation for national-level coordination, and the shifting governments in Portugal lacked a coherent transition policy (Minter 1994: 94). Rumours and deliberate anti-independence propaganda fostered white exodus in 1974-5. Ninety per cent of the Portuguese fled Mozambique, and with them went much experience in industry, commerce and government (Hanlon 1984: 188-9). Those Portuguese who stayed often found themselves in better positions than would have been possible for them in Portugal, even those with little education securing high posts in the tumultuous state of affairs that prevailed immediately after independence.

Revolution, the Cold War and apartheid The bias of governments in Western Europe and the United States towards Lisbon during the liberation struggle made a socialist alternative all the more natural for Frelimo's leadership. Already in 1968 the broadly anticolonial movement adopted Marxist perspectives on its struggle (Ottaway 1988: 213; Young 1988: 167). In 1977, Frelimo declared itself a MarxistLeninist vanguard party (Isaacman and Isaacman 1983: 121-2; Ottaway 1988: 214-15). According to the Frelimo leadership, Marxism-Leninism, in contrast with earlier experiments with 'African socialism' in some countries, was a revolutionary theory waiting to be applied 'creatively' in the

INTRODUCTION

7

Mozambican context (Munslow 1983: 155; de Brito 1988; Simpson 1993: 312-16). In practice, however, Frelimo's 'democratic centralism' (Hanlon 1984: 145) followed the oft-trodden paths of 'socialist' transformation. Plantations and factories, even village shops, were eventually nationalised. But even more significant for the subsequent criticism of Frelimo were the communal modes of production and administration, 'communal villages' (aldeias communais) in the countryside and 'communal neighbourhoods' (bairros communais) in small towns. Despite Machel's confident declaration that the communal village represented the 'decisive factor for the victory of socialism in our country' (quoted in Hanlon 1984: 138), the actual reach of the project was modest. Communal villages received little funding from the state and had incorporated only about 18 per cent of the rural population by the early 1980s (Hanlon 1984: 122). Moreover, communal villages with functioning cooperatives were in the minority, and, by 1983, 60 per cent of the communal villages were in the former 'liberated zones' (Isaacman and Isaacman 1983: 155). In some areas, such as in Tete Province, communal villages were often a direct continuation of the Portuguese resettlement schemes before independence (Borges Coelho 1993). The Frelimo leadership was anxious to ensure popular participation in its efforts to build a united nation-state. The tum to vanguardism replaced 'dynamising groups' with more hierarchical party structures of cells, circles, zones and districts, each headed by a secretary (Alexander 1997: 2-3). People's Assemblies were established to give non-activists a more audible voice in the management of districts and provinces. In practice, no policymaking or legislative functions accompanied the elected assemblies (Hanlon 1984: 144). Not only did the party set the policy but its leaders also prepared the lists from which assembly deputies were nominated. Many deputies were, in fact, active in Frelimo or its 'mass organisations' for women and youth. At no level were party and state clearly distinguished; the highest-ranking administrator at the provincial level was also the first secretary of Frelimo and the president of the assembly - the same pattern was replicated at the district level. The directors of state enterprises and farms were also often recruited from the ranks of party leaders. The age of multipartyism and decentralisation easily gives hindsight to bemoan Frelimo's early practices, but it should not eclipse the widespread hope and enthusiasm that characterised the first few years of independence. Not only was there a stream of cooperantes, foreigners sent by solidarity groups in Western Europe and by governments in Eastern Europe and Cuba, to ease the shortage of teachers, doctors and engineers in the country. Even more significantly, a sense of purpose and hope lifted the spirits of most Mozambicans themselves (cf. Minter 1994: 16). The question is, as mentioned, the extent to which Frelimo's project helped the subsequent

8

FROM WAR TO PEACE

tragedy to unfold. Clearly, the centralist solutions described above hardly sustained a sense of popular participation. The Frelimo party-state was, in effect, based as much on exclusion as on encompassing party structures. Highly significant contingents of the populace were excluded from the quintessential party membership. The reason was 'obscurantism', thought to be no less an enemy of the revolution than capitalist and feudalist social relations, illiteracy and ignorance (Simpson 1993: 325). Obscurantism included a range of 'traditional' beliefs and practices, from polygamy and the belief in the existence of spirits to the authority of chiefs and village headmen. Its agents were unable to join the party, whilst chiefs and headmen were deposed from their positions of authority. The Cold War made its impact not only through the havoc caused by the war but also through the conditions that Western governments, particularly the United States, set on their foreign aid. Large parts of Mozambique suffered from drought in 1983, and the final estimate of people starved to death was 100,000 (Minter 1994: 167). The Mozambican government had repeatedly appealed for relief aid, but food aid was actually reduced during the first half of 1983. This was the period when the government engaged in negotiations with South Africa to end its assistance for Renamo and Frelimo's support for the African National Congress. Such an agreement, however, was not the only condition which the United States imposed for Mozambique to receive its aid. It also wanted Mozambique to join the World Bank and the International Monetary Fund. Furthermore, it wanted relief aid to be distributed by the US and other foreign agencies, despite the Mozambican government's own system of relief distribution, at that time virtually free from corruption (Hanlon 1996: 16-17). Another stoppage in food aid took place in 1986, when Western donors urged Mozambique to accept its first Structural Adjustment package. Ever since, Mozambique has been one of the most foreign aid-dependent countries in the world. In 1990-4 it was the largest aid recipient in subSaharan Africa, with a total aid bill of over US$1.1 billion per year. Economic reforms, shaped by the politics of aid, thus preceded reforms in ideology and constitution, but they were enough to make Frelimo acceptable to Western donors (see Hanlon 1991). The US government's involvement in Mozambique's affairs did not extend to measures against Renamo's material support from groups and individuals in the US (Austin 1994; Minter 1994). A report commissioned by the US government documented Renamo's violent recruitment methods and atrocities (see Gersony 1988). But the government's reaction was merely to denounce Renamo, without pursuing the link between Renamo and its supporters in the US and South Africa. Assistance for Renamo by the South African military had been preceded by support from the Rhodesian government, which, according to the testimony of its intelligence officer

INTRODUCTION

9

(see Flower 1987: 300-2), virtually created Renamo by training and arming its guerrillas. Rhodesia and South Mrica had in Frelimo's Mozambique a common enemy, which was providing a safe haven for anti-apartheid liberation movements. After Zimbabwe's independence in 1980, South Mrica's economic interests also became apparent. The wars in Angola and Mozambique increased the dependence of such land-locked countries as Zaire, Zambia, Zimbabwe, Malawi, Botswana and Swaziland on South Mrican ports (Minter 1994: 117-20). As a post-independence resistance movement, however, Renamo did not simply appear from outside Mozambique's boundaries. As mentioned, movements other than Frelimo had fought colonialism, particularly in Zambezia Province, without challenging Frelimo as the largest nationalist movement. One of them was the National Union for the Liberation of Rombezia (UNAR), which appeared in 1968. Its objective was to create an independent territory between the Tanzanian border, the Rovuma river and the Zambezi river (Borges Coelho and Vines 1995: 32). The notion of Rombezia appealed to defectors from other nationalist movements who were looking for new ways to acquire political influence. It appealed, also, to President Kamuzu Banda of Malawi, whose agenda was to incorporate northern Mozambique into his vision of Great Malawi (see Chiume 1975: 167; Phiri 1985: 118-19). The most important supporter was Jorge Jardim, an influential businessman in Beira. With his aid, a new movement, Africa Livre, began operations in western Zambezia in 1976, launching attacks from Malawi. By 1982, after Jardim's death, Gimo Phiri led the movement to a merger with Renamo (Vines 1991: 53-8; Legrand 1993: 88-91). 2 Africa Livre is one example of Renamo successfully appropriating local discontent. During a drought in 1979, small groups of armed men already operated from Rhodesia and brought food and clothes to villages in central Mozambique, an area in which Frelimo had not yet established an effective presence (Hanlon 1984: 229-31). The central Mozambican provinces of Sofala and Manica became the initial settings for sustained Renamo presence. The Gorongosa mountains became the headquarters and Ndau the lingua franca of the movement. By 1983, the English-based acronym MNR had been abandoned and a network of semi-permanent bases in Mozambique had been established (Hall 1990: 40; Minter 1994: 41-2). Then began Renamo's expansion, and virtually the whole country became affected by the war.

From a civil war to peace Allen Isaacman, an eminent historian and Frelimo's early foreign sympathiser, dismissed Renamo as bandits, 'South African backed terrorists' (1988: 12). His verdict was dated as late as August 1987 (1988: 14), and a similar sentiment prevailed among many top government officials at least until the

10

FROM WAR TO PEACE

early 1990s. There was certainly evidence to support such a sentiment. The use of coercion in recruiting civilians to act as porters and eventually as guerrillas, and the· use of boy soldiers, were publicised in the most shocking details of early refugee accounts (see Gersony 1988; Magaia 1988; Minter 1989; Boothby et al. 1991). Academic analyses followed, pinpointing ritualised violence and terror as means both to initiate civilians into the movement and to demonstrate the movement's superior command of spirits and medical substances (see Nordstrom 1992; Wilson 1992a). These observations, together with the fact that Renamo's political agenda remained vague throughout the war (Geffray 1990: 41; Hall1990: 59; Finnegan 1992: 745; Alden and Simpson 1993: 124), could only underline destruction and terror as the movement's essence. Indeed, such was its meticulous destruction ofhealth and education services that it appeared to 'symbolically banish the state' (Wilson 1992a: 540; cf. Harilon 1991: 38-41; Vines 1991: 90). In actual fact, however, Renamo could appear in very different guises in different parts of the country and at different points in time. Extreme violence was by no means the general rule throughout Mozambique (Minter 1994: 182). An early ·report identified three different types of Renamo operations: the first in regard to taxation, the second in regard to control, and the third in regard to destruction (Gersony 1988: 10-20). Although somewhat rigid, this division has the merit of disclosing a pattern beyond destruction and terror. Renamo committed its worst atrocities, such as burning alive busloads of civilians in road ambushes, in Frelimo's southern strongholds (Hall 1990: 53). In parts of central and northern Mozambique, Renamo enjoyed some popular support and reinstated chiefs, healers and other agents of 'obscurantism' in their positions of authority (Geffray 1990; Wilson 1992a).3 The 1994 general elections, marking the end of the war, had an outcome which seemed to confirm this regional pattern of support. Renamo won most of the votes in many provinces throughout central and northern Mozambique, and the national results gave 38 per cent of the votes to Renamo and 44 per cent to Frelimo (Agenda de informa